Compatibility: Difficult and Easy Crosses

Most of the articles on this site focus on chromosome configurations and their implications for fertility. For example, crossing a tetraploid aril with Iris pumila is expected to produce seedlings that are fertile amphidiploids. But that invokes a different question, how difficult is it to make such a cross in the first place?

I find that some people confuse the issue of the difficulty of the cross with the expected fertility of the seedlings. They are not at all the same.

When judging the fertility of an iris, a key consideration is that to produce viable gametes (ovules or pollen grains), chromosomes must pair during the process of meiosis, a special kind of cell division that occurs only during gamete formation. If all the chromosomes pair readily, many viable gametes will be formed. If pairing fails or is incomplete, there may be no viable gametes or only a few.

To make a cross and produce seeds obviously requires that the pod parent have viable ovules and the pollen parent have viable pollen. If both these things are true, there is at least the potential for successful fertilization and seed production. Nevertheless, some combinations of parents are less likely to "take" and produce seeds than others. As an extreme example, you can take pollen from a perfectly fertile TB iris and apply it to a perfectly fertile petunia and never get any seeds. The two plants are simply incompatible. They are too different from each other to mate. If this were not the case, there would be no distinct species at all and evolution as we know it could not exist. Plants have a number of physiological mechanisms to prevent fertilization from outside their own species or group.

So one obvious barrier to fertilization is how closely related the two parents are. Similar parents cross more readily than dissimilar ones. Additionally, it is widely recognized that fertilization is difficult between plants of different ploidy, even if they are otherwise closely related. Crossing a diploid TB with a tetraploid TB is more difficult than crossing two tetraploids. Although there are some general rules like this, it is a somewhat quirky business in practice. Experienced hybridizers are aware that some combinations of parents just do not work out, even if the two irises concerned are both fertile with other irises of the same type.

Over the last five years, I have made hundreds of crosses between the different types of irises I work with, many of them with dozens of individual pollinations. There is enough data now to start to see some patterns in what kind of crosses readily take and produce seeds and what kinds are more problematic. I track not only the number of successful pollinations, but the total number of seeds produced. Sometimes a difficult cross succeeds, but only one or two seeds result. Seeds per pollination attempt seems a good measure of compatibility. Here are the data for the types of crosses I make often enough to get reasonable averages.

Diploid Aril X Diploid Aril AA X AA

number of pollinations

takes

takes/pollination (%)

seeds produced

seeds/pollination

22
7
32%
161
7.3

Diploid Aril X Tetraploid Aril AA X AAAA

number of pollinations

takes

takes/pollination (%)

seeds produced

seeds/pollination

25
6
24%
14
0.6

48-chromosome Bearded X Tetraploid Aril (or reverse) TTTT X AAAA

number of pollinations

takes

takes/pollination (%)

seeds produced

seeds/pollination

311
31
10%
209
0.7

Amphidiploid Arilbred X Diploid Aril AATT X AA

number of pollinations

takes

takes/pollination (%)

seeds produced

seeds/pollination

26
2
8%
11
0.4

Iris pumilaIris pumila PPPP X PPPP

number of pollinations

takes

takes/pollination (%)

seeds produced

seeds/pollination

235
51
22%
1995
8.5

Tetraploid Aril X Iris pumila (or reverse) AAAA X PPPP

number of pollinations

takes

takes/pollination (%)

seeds produced

seeds/pollination

269
60
22%
429
1.6

48-chromosome Bearded X Iris pumila TTTT X PPPP

number of pollinations

takes

takes/pollination (%)

seeds produced

seeds/pollination

56
12
21%
115
2.1

Amphidiploid (40-chromomsome) Bearded X Amphidiploid (40-chromomsome) Bearded PPTT X PPTT

number of pollinations

takes

takes/pollination (%)

seeds produced

seeds/pollination

34
11
32%
219
6.4

48-chromosome Bearded X 48-chromosome Bearded TTTT X TTTT

number of pollinations

takes

takes/pollination (%)

seeds produced

seeds/pollination

75
24
32%
560
7.6

I find it interesting that these results sort into two fairly distinct categories: easy crosses with 5-10 seeds per pollination, and difficult crosses with 0-2 seeds per pollination. All the easy types in my data set happen to be cases where the two parents are of the same type, there is a long history of experience to show that crosses between amphidiploids and one of the parent types (say, TB X SDB or TB X AB) are also relatively easy. I had surprisingly good success with the few AB X I. pumila crosses I have made, but there are not enough data to draw much of a conclusion from that.

It also occurs to me that advanced generation garden hybrids, such as TBs, may generally have wider compatibility than species, as there is a selection effect in favor of those which are less picky in responding to the hybridizers' suggestions.

As someone who is interested in expanding the fertile families through these wide crosses, I have sometimes wondered why there is not more interest in such projects. Part of the answer probably lies in these numbers. It takes about 10 times the effort per seed obtained.

 

Tom Waters

March 2017

More Articles

Return to Telperion Oasis

Unless otherwise noted, all text and illustrations copyright Tom Waters and all photographs copyright Tom or Karen Waters. Please do not reproduce without permission.